PI-5 sensitivity

[macross]

I have just finished reading ITMD second edition and I must say, this book is a gold mine of information for a beginner like me, thank you guys for taking the time to share your knowledge with people like me.

On page 167 (PI-4 design) it was mentioned that sensitivity of that particular design would be lower than the previous ones, because of direct sampling and a low resolution 10 bit ADC. After reading trough PI-5 design I started wondering, what if PI-5 design would be implemented using MCU and we would measure decay time using it's comparator instead of ADC(same as PICKINIV4), would sensitivity be greater than that of PI-4 design?

[geoscash1]

I would love to see someone get this circuit working. I'm with ya and will follow your progress if you go for the challenge. I have built it 4 times but to no avail yet.

I am working on giving it another try.

[bernte_one]

there must be something basicly wrong in this concept

[Qiaozhi]

there must be something basicly wrong in this concept

Actually, I was looking at this a few days ago, and I believe I know where the problem lies.

After examining the original prototype unit, I discovered that unfortunately there's a couple of errors in the schematic.
Please see the attached LTSpice simulation.

The original circuit was slightly different, in as much as it used two separate battery packs to achieve the +/- supplies, and an NMOS to switch the coil. The design in the book creates the two supplies using a phantom ground, and a PMOS to do the switching. The main issue is, however, that the 555 should be triggered (via C5) from the TX oscillator output, and not the preamp output. Also, due to the TX oscillator running from the same power rail (the original was on the other rail) you need to remove one of the NAND gates.The changes should be obvious from the LTSpice circuit.

[RufusGartz]

I have been looking at Qiaozhi's PI-5 MD Book 2.asc in LTSpice and the circuit in ITMD and at the original prototype circuit (https://www.geotech1.com/forums/show...400#post196400).

I replaced the V2 oscillator with a 555 timer oscillator plus nand gate that gives almost the same timings as shown in the LTSpice circuit in previous post. Original Tperiod = 1.5mS, Pulse Ton = 82uS. The theoretical timing calculation for the 555 oscillator plus nand gate in the attached circuit give Tperiod = 1.5mS and Pulse = 81.54uS. All seems to be working in LTSpice.

I have redone a jpg of the circuit to be the same as the LTSpice simulation.

Is there a way in LTSpice to simulate a piece of metal near the coil in the circuit to show the difference in audio output? Probably already been done before but I am after a link to the thread.

[Qiaozhi]

Is there a way in LTSpice to simulate a piece of metal near the coil in the circuit to show the difference in audio output? Probably already been done before but I am after a link to the thread.

You can use an inductor and resistor in parallel to represent a target. The time constant will be L*R, so (for instance) an inductance of 100uH with a 1 ohm resistor will have a tau of 100us. Then define a suitable coupling coefficient between the TX coil and the target inductance. Example: K1 L1 L2 0.1
Don't forget to connect a ground symbol to the target.

[RufusGartz]

Thanks for the help.

The circuit works. The sound gets louder near metal. Had to change the value of the cap to 47nF on the input to the trigger of the monostable 555 timer circuit. It didn't work with a higher value of 100nF. The blanking pulse width resistor definitely needs to be adjustable - 10K.

This metal detector would appear to make a good pin-pointer.

I've attached the updated files. (To see how the PI-5 works without metal near the coil comment out the inductor coupling directive.)

[bernte_one]

someone able to draw a pcb layout, inluding all corrections?